Approach to predictability via anticipated synchronization

Year: 2005

Authors: Ciszak M., Gutiérrez J.M., Cofiño A.S., Mirasso C., Toral R., Pesquera L., Ortín S.

Autors Affiliation: Department of Physics, University of Balearic Islands, E-07122 Palma de Mallorca, Spain; Department of Applied Mathematics, University of Cantabria, E-39005 Santander, Spain; Mediterranean Institute for Advanced Studies (IMEDEA), E-07122 Palma de Mallorca, Spain; Instituto de Física de Cantabria (CSIC-UC), E-39005 Santander, Spain; Departamento de Física Moderna, University of Cantabria, E-39005 Santander, Spain

Abstract: Predictability of chaotic systems is limited, in addition to the precision of the knowledge of the initial conditions, by the error of the models used to extract the nonlinear dynamics from the time series. In this paper, we analyze the predictions obtained from the anticipated synchronization scheme using a chain of slave neural network approximate replicas of the master system. We compare the maximum prediction horizons obtained with those attainable using standard prediction techniques.


Volume: 72 (4)      Pages from: 046218-1  to: 046218-8

KeyWords: Chaotic systems; Error of the models; Master system; Prediction techniques, Dynamics; Error analysis; Mathematical models; Neural networks; Synchronization, Chaos theory
DOI: 10.1103/PhysRevE.72.046218